Anthrax lethal toxin (LT) and edema toxin (ET) are made of three protein components: protective antigen (PA), lethal factor (LF) and edema factor (EF). PA is the cell binding moiety which binds to cellular receptors Tumor Endothelial Marker-8 (TEM8) and Capillary Morphogenesis Protein-2 (CMG2). After binding these cellular receptors, PA is proteolytically cleaved by ubiquitously expressed cell surface proteases, and forms an active oligomer which allows binding of the two enzymatic moieties, LF and EF and their transport into the cell cytosol. LF leaves several mitogen activated protein kinase kinases (MEKs) and rodent inflammasome sensor Nlrp1. EF is an adenylate cyclase which converts ATP to cAMP. These two toxins are the major virulence determinants of anthrax, so vaccine and therapeutic development against this disease primarily targets these proteins. Re-engineering of PA to selectively be activated by tumor cells also allows an opportunity for development of anti-cancer agents. In the last budget period we continued our work in identification and characterization of variable domains of camelid heavy chain-only antibodies (VHHs) against anthrax toxins. These proteins are small single chain, heat- and pH- stable entities that can access epitopes that conventional antibodies cannot. Furthermore, they can be produced in recombinant form and purified rapidly once their sequence is known. Fifteen VHHs with subnanomolar affinity for LF and EF were obtained from immunized alpacas and screened for anthrax neutralizing activity. Multiple classes of neutralizing VHHs were identified, including VHHs cross reactive against both toxins and preventing toxin entry into cells. Individual VHHs as well as bispecific neutralizing agents made from two linked neutralizing VHHs were tested in lethal anthrax spore infection models as well as mouse footpad edema models with success. In a separate related study, gene therapy with an adenoviral vector expressing a bispecific VHH-based neutralizing agent consisting of two linked VHHs targeting different PA-neutralizing epitopes was tested in two inbred mouse strains, BALB/cJ and C57BL/6J, and found to protect mice against anthrax toxin challenge and anthrax spore infection. This study demonstrated the potential for genetic delivery of antibodies as an effective method for prophylactic protection from anthrax. We also extended our studies with tumor-targeted anthrax lethal toxin to demonstrate that tumor cells which are normally resistant to anthrax toxins in vitro are highly sensitive to toxins when implanted in mice and that the host TEM8 and CMG2 receptors are required for toxin efficacy against these solid tumors. Using mice in which anthrax toxin receptors were selectively knocked out or exclusively expressed in various stromal compartments, we demonstrated that tumor sensitivity to toxin therapy requires CMG2 expression in host-derived tumor endothelial cells. Toxin inhibits proliferation of these cells through inhibition of the MEK pathways. Thus, the impact of the toxin therapeutics on tumor vessels and angiogenesis is the primary mechanism of action against solid tumors. In the same studies, we show that suppression of the host immune response with a regime of pentostatin and cyclophosphamide to overcome the immunogenicity of the toxin allows for an increase in the number of rounds of toxin therapy, and a significant improvement in therapeutic potential. In a related study on development of the anthrax toxins as anti-cancer agents, we used a phage display selection method to isolate PA variants with selective enhanced binding to CMG2 relative to TEM8. PA isoleucine residue 656 when mutated to valine or glutamine was found to have significantly reduced activity against TEM8-expressing cells while maintaining activity against CMG2-expressing cells. The mutant PA preference for CMG2 over TEM8 was also demonstrated in CMG2 and TEM8 knockout mice. We improved the specificity of previously constructed toxin anti-cancer agents in other studies. LF binding sites which require three subsites on two adjacent PA molecules were mutated in urokinase-activated and matrix metalloproteinase-activated PA variants in order to achieve higher specificity by requiring intermolecular complementation of these tumor-cell protease dependent PA molecules for inducing toxicity. Screening of a library of PA variants allowed selection of two optimal low activity mutants which when used in combination provided potent anti-tumor activity. We also collaborated on two studies related to anthrax toxin targeting of cancer cells. In the first, the toxicity of the urokinase-activated anthrax toxin against acute myeloid leukemia (AML) lines was correlated with urokinase plasminogen activator (uPAR) expression and MEK activity of each cell line. Bone marrow blasts and peripheral blood mononuclear cells were found to lack uPAR expression and were resistant to the toxin, demonstrating tumor selectivity in targeting of AML. In a separate study, a PA variant in which the sequence derived from protein C inhibitor replaced the native PA cleavage sequence was created. This variant was cleaved by the tumor associated serine protease testisin and showed increased killing of testisin-expressing cancer cells both in vitro and in vivo.
Showing the most recent 10 out of 44 publications
